Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
SPIE Advanced Lithography 2011 paper 7970-1<br />
e-<strong>Beam</strong> <strong>Direct</strong> <strong>Write</strong> <strong>of</strong> <strong>Wafers</strong><br />
Aki Fujimura, D2S, Inc<br />
1
What you do is amazing!<br />
Making S-Curves fit a Geometric<br />
If you have small S-Curves, you have to stack them one after another<br />
If you are blessed with a big S-Curve, you get to ride it for a long time<br />
2
“EbDW is always 1-2 generations behind”<br />
But that means it HAS kept up with Moore’s Law… Just needs to catch up<br />
3
<strong>Direct</strong>-<strong>Write</strong> or Mask?<br />
Either Way, e-<strong>Beam</strong> <strong>Write</strong>s All Chips!<br />
GDSII<br />
Design for<br />
E-beam<br />
(DFEB)<br />
Optimizes design data for<br />
e-beam writing<br />
<strong>Write</strong> Wafer<br />
<strong>Write</strong> Mask<br />
Contacts/Cuts First<br />
Enables Lower Volume at Leading Nodes<br />
Any Volume<br />
Any Layer
Similar but different…<br />
Need Mask Wafer<br />
Size 4x 1x<br />
Accuracy Highest (MEEF) High (less source <strong>of</strong> error)<br />
Data volume Huge (SRAF) Huge x 10 (no OPC; whole wafer)<br />
<strong>Write</strong> Time 8-40 hours 1 ~ 100 WPH<br />
Writing Space Rectangle (efficient) Circle (inefficient)<br />
# passes 2, 4 (accuracy) 1 (write time)<br />
Characters Circles 1000’s useful<br />
Market now Solid but flat Zero but big potential<br />
NIL Mask Master is a blend <strong>of</strong> both
e<strong>Beam</strong> Speed-Up techniques<br />
Split <strong>Beam</strong> Character Proj. General<br />
DFEB = Data Compression<br />
Maximum flexibility in<br />
shape being written<br />
MSB<br />
DFEB = 10X Speed-up<br />
Maximum transfer <strong>of</strong><br />
energy for all shapes<br />
Multiple Source<br />
Clustering<br />
Source Current<br />
Resist Sensitivity<br />
Regular Design Rules<br />
Borodovsky Method<br />
Other Limits:<br />
Stage<br />
Data Transfer<br />
Calibration<br />
Settling Time …<br />
Drawing Courtesy <strong>of</strong> Advantest<br />
SEM Courtesy <strong>of</strong> e-Shuttle, Inc.<br />
6
Advantest MCC8<br />
Realistic with limited budget is to extend<br />
X 10 Cluster<br />
7
Vistec MSB<br />
Hybrid like Prius is the best way to transition<br />
8
IMS, KLA-T REBL, Mapper, Multibeam<br />
Take on the bigger S-Curve<br />
Multiple beams individually<br />
controlled by various means<br />
Innovation in stage, data pipeline<br />
and column taken as a system<br />
Expensive but worth it<br />
9
Design to Manufacturing collaboration<br />
Focus to succeed : The “Borodovsky method”<br />
Design for E-<strong>Beam</strong> (DFEB)<br />
10
ASELTA, CEA/Leti, D2S, EQUIcon,<br />
Fraunh<strong>of</strong>er, Technolution, TSMC<br />
S<strong>of</strong>tware help is essential for this resolution<br />
MB-MDP on 80nm L:S for mask<br />
Mapper data prep for 3.5nm resolution<br />
Even e<strong>Beam</strong> writing is hard at these nodes<br />
• MB-MDP<br />
• MPC<br />
• eRIF<br />
• EBPC<br />
11
Themes and Approaches<br />
EbDW : we can do this….<br />
• Monetize from the Existing Mask Market<br />
• D2S, IMS, SEMATECH, Vistec<br />
• “Current density advances will not be enough” – Tom Faure, IBM<br />
• EbDW for high volume with clusters<br />
• EbDW can also enable the long tail<br />
• Government funding is deserved<br />
• EbDW will help boost design starts<br />
• That’s good for everyone in the supply chain<br />
12
Demand<br />
Enabling the Long Tail <strong>of</strong> SoCs<br />
The long tail<br />
Popularity Rank<br />
Source: Chris Anderson’s “The long tail: Why the future <strong>of</strong> business is selling less <strong>of</strong> more”
Revenue per Design<br />
The Tail is Getting Shorter<br />
We can enable the tail with EbDW w/DFEB<br />
More designs<br />
Faster time to market<br />
Cost <strong>of</strong> Manufacturing<br />
Chips per Design<br />
14-nm with mask<br />
20-nm with mask<br />
28-nm with mask<br />
Big opportunity<br />
Non-addressable Market<br />
Maskless SoC<br />
# <strong>of</strong> Designs
atio <strong>of</strong> revised to new designs<br />
Derivatives have low design cost<br />
10x reduction in mask cost increases derivatives by 10x<br />
100.0<br />
10.0<br />
1.0<br />
>=500-nm<br />
>=300 but =200 but =160 but =120 but =75 but
Electron <strong>Beam</strong> <strong>Direct</strong> <strong>Write</strong><br />
• Great for R&D : process, design, systems<br />
• Great for lower-volume : minimal mask cost<br />
• Volume Production? : 50 WPH cluster in 2014<br />
• Please attend these great sessions and learn<br />
about the potential <strong>of</strong> e-beam technologies
Session 3: Maskless Lithography I<br />
Date: Tuesday 1 March Time: 1:20 PM - 3:10 PM<br />
Session Chairs: Hans Loeschner, IMS Nan<strong>of</strong>abrication AG (Austria);<br />
Timothy R. Groves, Univ. at Albany<br />
E-beam lithography development, outlook, and critical challenges (Invited Paper)<br />
Paper 7970-9 Time: 1:20 PM - 1:50 PM<br />
Author(s): Hans C. Pfeiffer, HCP Consulting Services (United States)<br />
MCC8: throughput enhancement <strong>of</strong> EB direct writer<br />
Paper 7970-10 Time: 1:50 PM - 2:10 PM<br />
Author(s): Hideaki Komami, Masaki Kurokawa, Akio Yamada, Advantest Corp. (Japan)<br />
eMET: 50 keV electron multibeam mask exposure tool<br />
Paper 7970-11 Time: 2:10 PM - 2:30 PM<br />
Author(s): Christ<strong>of</strong> Klein, Jan Klikovits, Hans Loeschner, Elmar Platzgummer, IMS Nan<strong>of</strong>abrication AG (Austria)<br />
Scanning exposures with a MAPPER multibeam system<br />
Paper 7970-12 Time: 2:30 PM - 2:50 PM<br />
Author(s): Bert J. Kampherbeek, Christiaan van den Berg, Vincent Kuiper, Niels Vergeer, Stijn Bosschker, Thomas Ooms, Alexandra<br />
Tudorie, Remco J. Jager, MAPPER Lithography (Netherlands); Sjoerd Postma, DEMCON (Netherlands); Guido de Boer, MAPPER<br />
Lithography (Netherlands)<br />
Multishaped beam: development status and update on lithography results<br />
Paper 7970-13 Time: 2:50 PM - 3:10 PM<br />
Author(s): Ines A. Stolberg, Matthias Slodowski, Hans-Joachim Doering, Wolfgang H. Dorl, Vistec Electron <strong>Beam</strong> GmbH (Germany)<br />
17
Session 8: Maskless Lithography II<br />
Date: Wednesday 2 March Time: 3:30 PM - 5:40 PM<br />
Session Chairs: Lloyd C. Litt, SEMATECH Inc.; Laurent Pain, CEA-LETI (France)<br />
Fast mask writer: technology options and considerations (Invited Paper)<br />
Paper 7970-32 Time: 3:30 PM - 4:00 PM<br />
Author(s): Lloyd C. Litt, SEMATECH Inc. (United States); Timothy R. Groves, Univ. at Albany (United States); Gregory P. Hughes,<br />
SEMATECH, Inc. (United States)<br />
IMAGINE: an open consortium to boost maskless lithography take <strong>of</strong>f: first<br />
assessment results on MAPPER technology<br />
Paper 7970-33 Time: 4:00 PM - 4:20 PM<br />
Author(s): Laurent Pain, Serge V. Tedesco, Beatrice Icard, Mickael Martin, Christophe Constancias, Lab. d'Electronique de<br />
Technologie de l'Information (France); Bert J. Kampherbeek, MAPPER Lithography (Netherlands)<br />
Influence <strong>of</strong> massively parallel e-beam direct-write pixel size on electron proximity<br />
correction<br />
Paper 7970-34 Time: 4:20 PM - 4:40 PM<br />
Author(s): Shy-Jay Lin, Taiwan Semiconductor Manufacturing Co. Ltd. (Taiwan)<br />
Data path development for massive electron-beam maskless lithography<br />
Paper 7970-35 Time: 4:40 PM - 5:00 PM<br />
Author(s): Faruk Krecinic, Jack J. Chen, Shy-Jay Lin, Burn J. Lin, Taiwan Semiconductor Manufacturing Co. Ltd. (Taiwan)<br />
EBDW to complement optical lithography for 1D GDR patterning<br />
Paper 7970-36 Time: 5:00 PM - 5:20 PM<br />
Author(s): David K. Lam, David Liu, Multibeam Corp. (United States); Michael C. Smayling, Tela Innovations, Inc. (United States); Ted<br />
Prescop, Multibeam Corp. (United States)<br />
Model-based mask data preparation and impact on resist heating<br />
Paper 7970-37 Time: 5:20 PM - 5:40 PM<br />
Author(s): Aki Fujimura, D2S, Inc. (United States); Takashi Kamikubo, NuFlare Technology, Inc. (Japan); Ingo Bork, D2S, Inc. (United<br />
States<br />
18
Session 10: Maskless Lithography III<br />
Date: Thursday 3 March Time: 10:20 AM - 12:10 PM<br />
Session Chairs: Lloyd C. Litt, SEMATECH Inc.; Hans Loeschner, IMS Nan<strong>of</strong>abrication AG (Austria)<br />
New advances with REBL for maskless high-throughput EBDW<br />
lithography (Invited Paper)<br />
Paper 7970-43 Time: 10:20 AM - 10:50 AM<br />
Author(s): Paul Petric, Chris Bevis, Mark A. McCord, Allen Carroll, Alan D. Brodie, Upendra Ummethala, Luca<br />
Grella, Regina Freed, KLA-Tencor Corp. (United States)<br />
Large-scale eRIF implementation for sub-22-nm e-beam lithography<br />
Paper 7970-44 Time: 10:50 AM - 11:10 AM<br />
Author(s): Luc Martin, Lab. d'Electronique de Technologie de l'Information (France); Serdar Manakli, Sébastien<br />
Bayle, ASELTA Nanographics (France); Kang-Hoon Choi, Manuela S. Gutsch, Fraunh<strong>of</strong>er-Ctr. Nanoelektronische<br />
Technologien (Germany); Laurent Pain, Lab. d'Electronique de Technologie de l'Information (France)<br />
Demonstration <strong>of</strong> real-time pattern correction for high-throughput<br />
maskless lithography<br />
Paper 7970-45 Time: 11:10 AM - 11:30 AM<br />
Author(s): Marco J. Wieland, Ton van de Peut, Martijn Sanderse, MAPPER Lithography (Netherlands); Edwin<br />
Hakkennes, Nol Venema, Ard Wiersma, Mark Hoving, Sijmen Woutersen, Technolution B.V. (Netherlands)<br />
EBPC for multibeams low-kV electron projection lithography<br />
Paper 7970-46 Time: 11:30 AM - 11:50 AM<br />
Author(s): Jérôme Belledent, Sebastien Soulan, Laurent Pain, Commissariat à l'Énergie Atomique (France)<br />
Fast characterization <strong>of</strong> line-end shortening and application <strong>of</strong> novel LES<br />
correction algorithms in e-beam direct write<br />
Paper 7970-47 Time: 11:50 AM - 12:10 PM<br />
Author(s): Martin Freitag, Manuela S. Gutsch, Kang-Hoon Choi, Christoph K. Hohle, Fraunh<strong>of</strong>er-Ctr.<br />
Nanoelektronische Technologien (Germany); Michael Krüger, EQUIcon S<strong>of</strong>tware GmbH Jena (Germany); Ulf<br />
Weidenmüller, Vistec Electron <strong>Beam</strong> GmbH (Germany)<br />
19
It’s time to make it happen…<br />
• EbDW will scale, once it catches a node<br />
• Need to collaborate and coordinate even more<br />
• Too many teams in the league is diluting the quality and funding<br />
• Column, stage, data-path and integration should share<br />
• EbDW projects need more funding<br />
• Will help boost design starts at the leading edge nodes<br />
• Good for everyone in the supply chain<br />
• Spurs more silicon innovation<br />
• Fuels more systems innovation<br />
20